Beam-ACO Applied to Assembly Line Balancing

Abstract

Assembly line balancing concerns the design of assembly lines for the manufacturing of products. In this paper we consider the time and space constrained simple assembly line balancing problem with the objective of minimizing the number of necessary work stations. This problem is denoted by TSALBP-1 in the literature. For tackling this problem we propose a Beam-ACO approach, which is an algorithm that results from hybridizing ant colony optimization with beam search. The experimental results show that our algorithm is a state-of-the-art metaheuristic for this problem.

Multiobjective constructive heuristics for the 1/3 variant of the time and space assembly line balancing problem: ACO and random greedy search

“…Bautista and Pereira [5] proposed an ACO algorithm to solve a single-objective variant of the TSALBP, TSALBP-1, which tries to minimise the number of stations m, while fixing both the cycle time c and the station area A. That proposal is based on two previous papers that are applied to the SALBP [4,8], where the authors used a priority rules procedure with an ACO and a Beam-ACO algorithm, respectively. The latter proposal was later extended in [7].…”

“…Since the initial works of Dorigo et al [24], several researchers have developed different ACO algorithms that performed well when solving combinatorial problems such as the travelling salesman problem, the quadratic assignment problem, the resource allocation problem, telecommunication routing, production scheduling, vehicle routing, and machine learning [25,22,18,50,27,40,9]. Even the SALBP [4,8,7] and a single-objective variant of the TSALBP [5] have been solved by means of this kind of metaheuristic.…”

WoLF Ant

“…An early example of this include the candidate lists employed by ACO [1], [12], [13]. Recent approaches include beam searches [36], [37] and constraint checking [27]. Alternatively, using hierarchies, the search space is limited and approximated [22].…”

Swarm Intelligence in Optimization

“…Recent applications of ACO to problems arising in these areas include the applications to protein folding [153,154], to multiple sequence alignment [127], to DNA sequencing by hybridization [20], and to the prediction of major histocompatibility complex (MHC) class II binders [86]. ACO algorithms are currently among the state-of-the-art methods for solving, for example, the sequential ordering problem [62], the resource constraint project scheduling problem [120], the open shop scheduling problem [14], assembly line balancing [15], and the 2D and 3D hydrophobic polar protein folding problem [154]. In Table 2 we provide a list of representative ACO applications.…”

“…Furthermore, BS algorithms reduce the search space in the hope of not excluding all optimal solutions, while ACO algorithms consider the whole search space. Based on these observations Blum introduced a hybrid between ACO and BS which was labelled Beam-ACO [14,15]. Beam-ACO is an ACO algorithm in which the standard ACO solution construction mechanism is replaced by a probabilistic beam search procedure.…”